swagger-analyzer

Analyzes swagger (OpenAPI) specification.

Table of Contents

• What is it?
• Why I have to use it?
  • If you don't want
  • If you do want
• Installation (Coming soon)
• Usage (Coming soon)
• Examples (Coming soon)
• Changelog
• How to Contribute
• How to Make Pull Request
• License

What is it?

It is library created by huge fan of Swagger (OpenAPI) Specification.
I believe that this spec is not only about documenting projects but has a great potential in code generating, which will save a lot of time for developers.
This library will provide an easy and highly extensible way for creating high quality code-generators using next features:

• grouping operations for controllers on Backend(BE) and services on Front-end(FE) in order to avoid generating one fat file for API client:
  • by fist tag in it;
  • by convention for operation Id in it;
  • by your custom selector;
• grouping definitions in namespaces for both BE and FE:
  • by convention for its name;
  • by its appearance in operations(if they are grouped);
  • by your custom selector;
• detecting inline enums + their duplicates and inserting them in namespaces with definitions(if they are grouped):
  • by convention for its name;
  • by its appearance in operations(if they are grouped);
  • by its appearance in definitions(if they are grouped);
  • by its appearance in operations and definitions;
  • by your custom selector;
• grouping operations, definitions and inline enums in components:
  • by operations groups;
  • by definitions namespaces;
  • by your custom selector;
• using different strategies for grouping in components:
  • with putting all related objects in component in order to make it self-contained (so generated code will be more safe to edit);
  • with defining some objects as dependencies if it is present in more than 1 component (so generated code will be more clean);
  • with your custom strategy;
• generating objects for further BE mocking with human-readable response:
  • for response object:
    • by using its examples;
    • by using convention for its examples;
  • for schema object:
    • by using properties' format;
    • by using properties' examples;
    • by using convention for properties' examples;
    • by detecting patterns in properties' name;
  • by your custom strategies for: paths, operations, response object and schema object;
  • by combination of any number of strategies;
• supporting vendor extensions;
• supporting custom parsers that enables specific vendor extensions;
• supporting custom parsers that enables OpenAPI 3.0 (next) features:
  • parser for enabling schema and $ref in all type of parameters see next spec;
  • your custom parser.

Why I have to use it?

As for now you can't use it, because it isn't even published to NPM yet.
This repo exists only for that who want to participate in creating of this project.

If you don't want

Please return here at 2016-09-01 and try:

• This library for creating your own high quality code-generators;
• CLI tool inspired by Yeoman for integrating it in development process;
• Client code + mocks generator for Angular 1 + TypeScript;
• Client code + mocks generator for Angular 2;
• And may be something more (like plugins for Grunt, Gulp).

If you do want

First of all - any, absolutely any, help will be highly appreciated.
Second - if you want to help with coding read CONTRIBUTING.md. There are instructions to build, run and test it.
And last - some history and plans:
I'm using swagger for code generating nearly a year. I've tried dozens of tools based on swagger, but result wasn't worth it. In the end I've written small library and grunt wrapper for it.
Continuously improving them to fit needs of two my main web projects (one of them is already released) I've get inspiring results:

• It works :D
• Generated services for Angular 1.5 + TypeScript saves many hours for me and other FE developers in team;
• Grouping services, models and enums in components with strong typing gives really clean and maintainable project structure;
• Support for schema.$ref in all type of parameters from next spec makes code even better;
• Generating mocks with detecting patterns in properties' name in order to make human-readable responses for ngMockE2E brings possibility for user test of FE with nearly real data with minimum coding even if BE isn't implemented at all;
• Previous paragraph makes project really decomposed into BE and FE (except using websockets and chunk file uploading);
• And a lot of other small workflow improvements.

Unfortunately, while trying to add new features, I realized that, as it often happens, I was doing right things in wrong way.
After code review, I concluded that my code is too coupled and has some sort of bad practices in TypeScript and Node.js world.
So now I'm refactoring and porting it, because it has most of this features implemented.
Thanks for reading;)

Installation (Coming soon)

Usage (Coming soon)

Examples (Coming soon)

Changelog

Recent changes can be viewed on the CHANGELOG.md

How to Contribute

Read to contribute CONTRIBUTING.md

How to Make Pull Request

Read to contribute PULL_REQUEST_TEMPLATE.md

License

Copyright (c) Ihor Chulinda.
This source code is licensed under the MIT license.